What problem does it solve? Choosing which AI model each pstack role (critics, runners, investigators, workers) should use is tedious and error-prone when done by hand. This Skill detects the models actually available in your session and writes a single override rule so every pstack skill uses the right model per role. ## Core Features & Use Cases - Model Detection: Enumerates the model slugs available to Task subagents in the current session, never writing an unconfirmed slug. - Rule Generation: Writes ~/.cursor/rules/pstack-models.mdc with alwaysApply: true, mapping every pstack role to a chosen model, with inherit-parent and auto as always-valid aliases. - Validation & Idempotency: Validates every real slug against the detected set and overwrites the whole file so re-runs stay consistent. - Use Case: You want pstack's arena runners and critic panels to run on specific models you have access to. Run this Skill, confirm the role-to-model mapping, and it writes the override rule that all pstack skills read on their next session. ## Quick Start Ask the assistant to run /setup-pstack to detect your available models and configure which model each pstack role uses.